This material is based upon work supported by USSTRATCOM through the 55th Contracting Squadron under Contract No. FA4600-12-D-9000 and was also funded, in part, by NSF-1638099 and USDA-2017-67021-25924.
Underground sensor installation:
Predicting digging success:
Unmanned Aircraft Systems (UAS) are commonly used for data collection in a wide range of applications. This is usually done by equipping the systems with on-board sensors such as cameras and laser scanners for remote sensing. Additionally, UASs can also be used to collect data from sensors previously deployed in the field. A less explored modality for UASs consists of utilizing the UAS to actually deploy sensors in locations that are challenging to reach by other means. This is particularly challenging when the sensor placement has special requirements in terms of position, location, and general manipulation.
In this project we develop a system and experimentally verify the performance of a UAS that can carry a sensor and insert it into the soil via an auger mechanism. Placement of sensors underground requires that the UAS is able to identify the proper target location, break the ground surface, remove the soil, and correctly place the sensor. A significant challenge is how to perform these tasks successfully within the weight and power constraints of a UAS.
An additional challenge is being able to quickly determine if a digging evolution will succeed or not. Due to the finite amount of energy available to system from its batteries and the fact that flying and digging consume the majority of that energy, rapidly deciding that sensor emplacement is not feasible will allow for UAS repositioning to another location where an additional digging attempt can be made. We have developed a first-of-its-kind decision making algorithm using a Markov decision process to solve this problem.
- Carrick Detweiler (PI) (Computer Science and Engineering)
- Sebastian Elbaum (Former Faculty) (Computer Science and Engineering)
- Benjamin Terry (Mechanical Engineering)
- Justin Bradley (Computer Science and Engineering)
- Brittany Duncan (Computer Science and Engineering)
- Adam Plowcha (Computer Science and Engineering)
- Yue Sun (Mechanical Engineering)
- Mike Turner (Computer Science and Engineering)
- Marc Lussier (Mechanical Engineering)
- Alisha Bevins (Computer Science and Engineering – ugrad)
- Mark Nail (Mechanical Engineering – ugrad)
- Research/Engineering Staff
- Jacob Hogberg (Computer Science and Engineering)
- Y. Sun, A. Plowcha, M. Nail, C. Detweiler, S. Elbaum, and B. Terry. “Unmanned Aerial Auger for Underground Sensor Installation.“ 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2018)
- A. Plowcha, Y. Sun, C. Detweiler, and J. Bradley. “Predicting Digging Success for Unmanned Aircraft System Sensor Emplacement.” 2018 International Symposium on Experimental Robotics (ISER 2018).